A PDF homepage with just 50 likes pushed a Vietnamese game company's cleaner app to 1.19M installs

AI Cleaner: Storage Optimize is one of roughly twenty Android photo-cleaner apps named "AI Cleaner." Its name and package follow the category leader, its features are industry standard, and it reached 1.19M installs in about two and a half months. Its rise has less to do with the product than with a Vietnamese game publisher behind it: an in-house paid acquisition team, a row of old homepages ready to run ads, and a developer account dedicated to tool apps. Indie developers can learn how a game company approaches utility apps from it, but building another cleaner app to copy it isn't worth it.

Sep 25: a PDF homepage starts running cleaner ads

On Facebook, a page called "PDF Reader - Document Scanner" with just 50 likes began running ads in October 2025 and has run 331 in total. In Sep 2026 25, it started promoting an app with nothing to do with PDFs: AI Cleaner: Storage Optimize. Meta's public ad library shows 32 ads tied to this app, all launched after 9-25. About sixty percent are video, running on Facebook, Instagram, Threads, and Audience Network. Targeted countries visible include the US, UK, Germany, and Japan. The copy comes in English, French, and Spanish, with one hook:

"Memory full? 😱 AI Cleaner deletes hidden cache & duplicate photos in 1 tap!"

The same day, a Google Play review appeared, later marked "helpful" by 236 people, one sentence long: "repetitive ads in the way of cleaning" (ads keep getting in the way of cleaning).

Before that, this app, which launched on Aug 24, had only put up a little over 30k installs in a month (AppstoreSpy third-party estimate, 09-24).

What it is

An Android photo cleaner: it finds similar, duplicate, and blurry photos and screenshots, finds duplicate videos, lets you swipe through your gallery to sort, compresses photos and videos, and adds app management, a system recycle bin, and battery info. The free version has ads and daily limits on deletions and compressions. Pro removes the limits and ads, with in-app prices listed in the store from $1.99 to $59.99.

The store copy is more thoughtful than its competitors'. Rivals like to write "your phone is almost full," while this one says "you'll see everything before any deletion," and when comparing similar photos you can tap "Why this photo?" to see the reasons: sharpness, whether eyes are open, resolution, file size. Starred photos are never deleted, analysis runs on the phone, and you can grant access to only some photos. The text reads as if it was written to match Google Play's policies.

The ads themselves still say "Memory full? 😱". The review pinned to the top: "did not work and then it asked for payment." (10-15, 398 people found it helpful). The developer replied the next day, asking users to email [email protected]. By Nov 6, the store page showed a rating of 3.6 to 3.7 stars, from about 1,290 ratings, roughly one per thousand installs.

Who's behind it

The developer account is named "AI Translator," which looks like a translation tool author. At the bottom of the store page under "About the developer," the contact is Le Minh Duong, email [email protected], address in Binh Thanh District, Ho Chi Minh City. The website is vnpublisher.com, which search results list as "BigThink Game Studio."

Falcon Games is a Vietnamese mobile game publisher. Its website says "500+ Games Launched, 3B+ Downloads." On Google Play, 30 apps with package names starting com.fc. turn up, scattered across developer accounts under FALCON GAMES, CYLINDER GLOBAL PTE. LTD., Moboco, QUEEN CITY, and AI Translator. The biggest is Jelly Master: Mukbang ASMR, with 72.3M installs (Google Play, 11-06). The link between the AI Translator account and Falcon is inferred from the email, the package prefix, and the website; there's no official confirmation (inference).

Under the AI Translator account is a row of tools: XTranslator with 770k installs, Tarova tarot with 730k, PDF Reader with 380k, Lockora app lock with 200k, Calora calorie counter with 27k, plus a video downloader, Vidora, and a TV remote, XRemote, listed in the store. The cleaner app is the newest to break out of this row.

How the 1.19M came about

(Before 10-04, figures are AppstoreSpy estimates; from 10-29 onward they are public Google Play data.)

The curve has three phases. From 9-25, Meta ads begin and volume starts climbing. From 10-02 to 10-04, it added about 340k over two days, then stayed nearly flat for a week. After mid-October it settled at 20k to 50k per day, and in the last 7 days it added 182k.

340k in two days followed by an abrupt stop doesn't match Meta's usual steady daily spend. It looks more like a concentrated budget burst: possibly Google App campaigns, or cross-promotion inside the company's games. Falcon's games add hundreds of thousands of new installs a week, and placing a cleaner-app ad inside a game costs almost nothing. There's no direct evidence for either (inference).

Two other things helped. First, the name: the category leader is iKame's "AI Cleaner: Clean up Storage" (package com.aicleaner.clean.cleanstorage.ai, launched in March, already 11M installs). This one is "AI Cleaner: Storage Optimize" (package com.fc.ai.aicleaner.clean.cleanstorage.optimize), so the two show up side by side when people search "AI cleaner". Second, the countries: in November 6 it hit #7 on the Hong Kong Tools Top Free chart (its best was #6), #47 overall in Hong Kong, #33 in the Korea Tools chart, #73 in Poland, and it ranked 100 in Taiwan, Singapore and Vietnam. The Tools charts in these small storefronts have low bars, so charting there picks up a bit of organic traffic.

Nobody is talking about it on TikTok, Reddit or X, and no media has covered it. The only results online are auto-generated listing pages on APKPure and AppBrain.

Who it's competing with

Searching "phone cleaner" in the Meta ad library, there are 16,358 ads running in the US store as of November 6. Roughly 20 Google Play apps have "AI Cleaner" in the title, most launched between April and September of 2026 (Google Play installs, 11-06):

Add in Files by Google, Google Photos' "free up space" feature, and the built-in device maintenance on Samsung phones. All free, no ads, and on every phone. Among the launches from the same period it sits in the middle-to-lower range. The two JTR Labs apps get three to four times its weekly volume.

Falcon is cloning it too: in October 26 it launched "CleanX: AI Storage Optimizer" under the FALCON GAME account (11-06 with only 3,313 installs, already showing up on the Hong Kong Tools chart). In October 28 the iOS version went live under CYLINDER GLOBAL, with no ratings yet.

Where the money comes from

Ads plus subscriptions. Revenue isn't public, and it has never charted in any country's Top Grossing, while iKame, TapSuite and Codeway's apps have. Subscription conversion is probably low, so ads likely carry the economics (inference). Paid acquisition has never stopped and they keep spinning up clones, which suggests the numbers work for the operators, but how much it actually earns is not visible from outside.

How many people know about it

Hardly anyone knows this app. But everyone knows the business: twenty-odd apps with the same name, over ten thousand cleaning ads running in the US store, and copying each other is simply normal for this category. The app itself is a copy too. "Nobody knows it" here doesn't mean there's an open slot.

If it were you

Grade C. It's worth understanding how game companies build tool apps sideways, but don't build another AI photo cleaner. This one got traction through the publisher's paid acquisition team, a stack of old ad pages, and traffic from their own games. You don't have those, and even if you did, twenty same-name apps are already in the way.

Who it suits: small game studios or tool portfolio teams that already have a paid acquisition team and ad monetization experience and want to move that capability into tools. For them the thing to take is the playbook, not this product.

Who it doesn't suit: first-time app makers relying on organic store traffic; people who haven't run paid ads and can't calculate what an install is worth; anyone who can't run a multi-account portfolio or can't afford account bans. Cleaner apps are among the categories Google Play polices most strictly, and anxiety-driven ads and paywall complaints invite review.

How to do it: borrow three things. One: run a portfolio of tools under one account: translator, PDF, app lock, tarot, calorie counter, cleaner, sharing one ad SDK and subscription framework, and put more money behind whichever app pays back on ads. Two: run new apps through old Facebook pages, since pages with a paid history get through review more smoothly, and test a batch of creatives in English, French and Spanish templates at once. Three: write store copy that is transparent (confirm before deleting, local processing, access only to selected photos). That copy is for the reviewers; ad creative is a different job. If you just want to build a cleaner, at least avoid the same name and the same features. What nobody does well in photo cleaning is the segment that doesn't want to be bothered by ads and paywalls. A one-time purchase, no-ads version might find users in the review sections (inference; demand size not verified).

How to acquire users: paid acquisition. On Meta, run "storage full" video creatives targeting the US, UK, Germany and Japan, or use Google App campaigns. If you have your own traffic, cross-promote it. Start by pushing for the top spot in low-bar Tools storefronts like Hong Kong, Korea and Poland to pick up some organic traffic. This category has no content angle, so don't expect TikTok to carry it.

When to get in: there's no real window. This category filled up two years ago. It's still growing by about 2 to 50k a day. When that stops, it will mean paid acquisition no longer pays off in this category either.

Signs you've misread it:

• If it charts on a country's Tools Top Grossing list, or if subscription prices and paywalls get tougher while installs keep rising, that shows subscriptions can earn money and the category is less thin than it looks. Worth re-running the numbers then.
• If its daily new installs fall to a few thousand and it drops out of the top 50 on the Hong Kong Tools chart, paid acquisition has stopped and the math didn't work. Then its playbook isn't worth copying.
• If it or a sibling app in the same family gets pulled from the store, that signals Google is cracking down on multi-account portfolios and name-copying.

Sources

• Installs, ratings and charts: public Google Play data, 2026-11-06. Historical cumulative installs before 10-04 are estimates from the third-party AppstoreSpy.
• Store listing copy, reviews, developer info: Google Play store page, 2026-11-06.
• Ads: Meta ad library (the "PDF Reader - Document Scanner" page; US-store counts for the "phone cleaner" keyword), 2026-11-06; admanage.ai ad library page.
• Company: falcongames.com and Unity's case study page on Falcon Games.
• iOS version: public App Store lookup, id6805403614.
• Content channels: X, TikTok, Reddit and media search, no relevant results in the last 60 days.